1
1
<p align =" center " >
2
- <img width =" 70% " height =" 70% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /boundary-3D.png " >
2
+ <img width =" 70% " height =" 70% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /boundary-3D.png " >
3
3
</p >
4
4
5
5
<h3 align =" center " >Support Vector Data Description (SVDD)</h3 >
31
31
32
32
## Notices
33
33
34
- - This version of the code is not compatible with the versions lower than R2016b.
34
+ - This version of the code is not compatible with the versions lower than *** R2016b*** .
35
35
- The label must be 1 for positive sample or -1 for negative sample.
36
36
- Detailed applications please see the demonstrations.
37
37
- This code is for reference only.
@@ -54,8 +54,8 @@ A class named ***DataSet*** is defined to generate and partition the 2D or 3D ba
54
54
[trainData, trainLabel, testData, testLabel] = DataSet.partition(data, label, 'type', 'hybrid');
55
55
```
56
56
<p align =" center " >
57
- <img width =" 30% " height =" 30% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /banana-2D.png " >
58
- <img width =" 30% " height =" 30% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /banana-3D.png " >
57
+ <img width =" 30% " height =" 30% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /banana-2D.png " >
58
+ <img width =" 30% " height =" 30% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /banana-3D.png " >
59
59
</p >
60
60
61
61
### 02. Kernel funcions
@@ -184,7 +184,7 @@ svplot = SvddVisualization();
184
184
svplot.ROC(svdd);
185
185
```
186
186
<p align =" center " >
187
- <img width =" 40% " height =" 40% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /ROC-3D.png " >
187
+ <img width =" 40% " height =" 40% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /ROC-3D.png " >
188
188
</p >
189
189
190
190
The decision boundaries (only supported for 2D/3D dataset) are
@@ -194,28 +194,28 @@ svplot = SvddVisualization();
194
194
svplot.boundary(svdd);
195
195
```
196
196
<p align =" center " >
197
- <img width =" 70% " height =" 70% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /boundary-2D.png " >
197
+ <img width =" 70% " height =" 70% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /boundary-2D.png " >
198
198
</p >
199
199
200
200
<p align =" center " >
201
- <img width =" 70% " height =" 70% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /boundary-3D.png " >
201
+ <img width =" 70% " height =" 70% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /boundary-3D.png " >
202
202
</p >
203
203
204
204
The distance between the test data and the hypersphere is
205
205
```
206
206
svplot.distance(svdd, results);
207
207
```
208
208
<p align =" center " >
209
- <img width =" 40% " height =" 40% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /distance-3D.png " >
209
+ <img width =" 40% " height =" 40% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /distance-3D.png " >
210
210
</p >
211
211
212
212
For the test results, the test data and decision boundary (only supported for 2D/3D dataset) are
213
213
```
214
214
svplot.testDataWithBoundary(svdd, results);
215
215
```
216
216
<p align =" center " >
217
- <img width =" 40% " height =" 40% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /boundary-tets-2D.png " >
218
- <img width =" 40% " height =" 40% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /boundary-tets-3D.png " >
217
+ <img width =" 40% " height =" 40% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /boundary-tets-2D.png " >
218
+ <img width =" 40% " height =" 40% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /boundary-tets-3D.png " >
219
219
</p >
220
220
221
221
### 05. Parameter Optimization
@@ -241,8 +241,8 @@ svddParameter = struct('cost', cost,...
241
241
242
242
The visualization of parameter optimization is
243
243
<p align =" center " >
244
- <img width =" 40% " height =" 40% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /bayesopt.png " >
245
- <img width =" 40% " height =" 40% " src =" https://github.com/iqiukp/Support-Vector-Data-Description-SVDD/blob/master/img /bayesopt-1.png " >
244
+ <img width =" 40% " height =" 40% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /bayesopt.png " >
245
+ <img width =" 40% " height =" 40% " src =" https://github-files-qiu.oss-cn-beijing.aliyuncs.com/SVDD-MATLAB /bayesopt-1.png " >
246
246
</p >
247
247
248
248
** Notice**
0 commit comments